Background and Purpose: Traumatic brain injury (TBI) has been described as a signature injury of the OIF-OEF-OND combat experience. Chronic memory deficits due to problems in memory encoding, consolidation, or retrieval, are often seen after moderate-severe TBI (M- STBI) and may pose barriers to community re-entry, retraining, and rehabilitation. Studies have demonstrated some efficacy in memory rehabilitation using cognitive treatments such as memory strategy training, errorless learning techniques, retrieval practice, and other methods. However, some patients do not benefit from such interventions, while others minimally improve. No prior group study has attempted to match the Veteran's specific information-processing deficit with a memory rehabilitation therapy targeted to their specific problem (patient-treatment matching), though such an approach should produce the best results. We propose to develop and test three memory therapies, each designed to foster improvement in a particular stage of memory processing. Veterans will then be classified by their primary locus of impairment and will participate in (a) a treatment matched to their deficit, and (b) a mismatched treatment. It is predicted that Veteran participants will show greater benefit from the matched treatment. If successful, this research could guide decision-making regarding the most efficacious approaches to alleviating disabling memory deficits in Veterans with M-STBI. Procedures: The proposed research takes place in three phases. In Phase 1, three memory therapies (memory strategy training, errorless learning, and retrieval practice) will be formalized through the development of a treatment manual. Therapists will be trained to provide these treatments with high fidelity, and their adherence to treatment regimens will be measured. In Phase 2, Veterans with chronic memory deficits from M-STBI will be evaluated with neuropsychological (NP) and neuroimaging (MRI) probes. The NP probes will allow us to classify participants into three groups, according to the primary locus of their memory deficit in encoding, consolidation, or retrieval. MRI sequences enable discovery of structural (morphometric and diffusion imaging) and functional (resting state functional connectivity) biomarkers of this classification, for use i a later trial evaluating the neural basis of treatment improvement. In Phase 3, Veteran participants will undergo treatment with two therapies, one matched to their deficit, the other mismatched. Treatment effect will be measured by assessing change on NP measures and on real-life functional measures and quality of life. Key Questions: Our primary goal is to develop and test the efficacy of three process-specific memory interventions in a heterogeneous population of Veterans with chronic memory impairment after M-STBI. We aim to learn whether improved efficacy results from appropriate patient-treatment matching. We also hope to refine a classification system that reliably identifies encoding, consolidation, and retrieval deficits in or participants, and to identify specific structural and functional brain biomarkers of this classification scheme. For our long term goal, we aim to maximize the efficacy of memory improvement therapies for alleviating memory impairment after acquired brain injury, and to use both NP and brain imaging biomarkers in better understanding mechanisms of change resulting from rehabilitative intervention. This knowledge will enhance discovery of best practices for treating debilitating memory dysfunction and maximizing functional recovery after injury.